How Nervous is your Supply Chain Planning System?

In a recent McKinsey report it was stated “After the implementation of a new planning and scheduling system, planners at one pharma manufacturer started receiving more than 200 exception messages every day.” This phenomenon is a consequence of a system which is not smart. In other words, the system fails to understand the relevance and the impact of each event on the supply chain operations. Systems should be capable of handling and automating responses to high frequency and low impact events. In addition, to augment and speed up decision making with high impact and low frequency events.

Unfortunately, most current supply chain technology is based on giving visibility without understanding what matters. Knowing and evaluating the impact of each event and having a solution requires a great deal of expertise. Consider a supplier sending a message indicating their late delivery by 2 days. Does it matter? Will it cause shut down of production or a late delivery to your end customers? What if the delay is 2 hours or 2 weeks? Enabling a system to interpret the messages as received and reacting as needed requires deep modeling capability of the supply chain as well as intelligence. For example, if the supply late arrival really matters, then can we possibly delay another order in favor of the one becoming late, determine if it is possible to use a substitute part instead, or perhaps pay the premium to get the part from another vendor?

Exploring such possibilities shows how the system can respond better. However, intelligent systems can predict many of such issues occurring and take preventive actions. In the above example, it could be a pattern of behavior from the supplier during a certain season that can be expected and taken into account. Another example could be a customer that places a large order at certain points in time causing such delays.  Intelligent systems enable you to predict such patterns ahead of time and mitigate supply chain risks providing much better commitment dates at a lower cost.

Finding patterns of behavior in the supply chain using machine learning techniques helps to make the supply chains more resilient and more predictable.

The more predictive you are the less reactive you need to be

Furthermore, having a distributed environment that uses intelligent sensors to detect, interpret and acts on events accordingly prevents sending hundreds of exception reports, creating nervousness of the system.  For more information on Adexa Intelligent and Distributed Genies click Here.

“Interpreting the messages as received and reacting as needed requires deep modeling capability of the supply chain as well as intelligence.”